Nagios Support Forum

So that is your problem, there is a bug in XI 5.3.0 that happens when running PHP < 5.2.x. The question becomes why didn't the installer upgrade script do this. In the directory you ran the upgrade script from there should be a file called upgrade.log(assuming you ran it from the CLI). Can you PM that to me as well?

Also what is your OS/Arch?

CentOS 5.11 x32. We have attempted to migrate over to a CentOS 7 x64 server, but ran into problems. Still hoping to get there at some point.

So the answer is to upgrade php for now?

Upgrading PHP on Cent 5 is going to be problematic and not in scope of support.

If it was me this is the process I would use. Assuming you have some type of hypervisor setup:

Make a complete backup. You don't have a backup unless you have a tested restore.
Make an XI only backup Backup XI independently
Install Cent 7 on new VM.
Install 5.2.9 and activate.
Restore XI only backup to new VM.
Upgrade from 5.2.9 to 5.3.0.
Shutdown old VM and change it's IP to new VM.

^ If you don't follow in order there can be serious repercussions.

We upgraded to 5.3.0 and are running into same issue with session time out. Upon checking php version we are also not running 5.2 or later. We are running Redhat 5.11. We upgraded from XI 2012R9. Are we not able to just upgrade php version manually?

Thanks

I have sent this issue to the developers for further analysis. If there is a better method of handling this, I will update this thread.

avandemore wrote:Upgrading PHP on Cent 5 is going to be problematic and not in scope of support.

If it was me this is the process I would use. Assuming you have some type of hypervisor setup:

Make a complete backup. You don't have a backup unless you have a tested restore.
Make an XI only backup Backup XI independently
Install Cent 7 on new VM.
Install 5.2.9 and activate.
Restore XI only backup to new VM.
Upgrade from 5.2.9 to 5.3.0.
Shutdown old VM and change it's IP to new VM.

^ If you don't follow in order there can be serious repercussions.

When you say 'Restore XI only backup' are you referring to only the Nagios XI files (/usr/local/nagiosxi)? Or following the procedure you linked?

Thanks

Or following the procedure you linked?

The backup archive created during the procedure.

I'd also like to clarify that we did not remove CentOS/RHEL 5.x support in 5.3.0 - there is a bug due to some functions that were used not being backwards compatible with PHP < 5.2 and we will resolve these issues in XI 5.3.1 - we've already filed a bug report internally for this issue.

avandemore wrote:Upgrading PHP on Cent 5 is going to be problematic and not in scope of support.

If it was me this is the process I would use. Assuming you have some type of hypervisor setup:

Make a complete backup. You don't have a backup unless you have a tested restore.
Make an XI only backup Backup XI independently
Install Cent 7 on new VM.
Install 5.2.9 and activate.
Restore XI only backup to new VM.
Upgrade from 5.2.9 to 5.3.0.
Shutdown old VM and change it's IP to new VM.

^ If you don't follow in order there can be serious repercussions.

I had a few problems after following the above that caused me to go back to the Cent 5 system - graphs not updating, some alerts having notifications disabled (and not being able to enable), and mail alerts continuously being sent (even after I cleaned all outgoing mail and no new alerts were coming into the system).

I had a few problems after following the above that caused me to go back to the Cent 5 system - graphs not updating, some alerts having notifications disabled (and not being able to enable), and mail alerts continuously being sent (even after I cleaned all outgoing mail and no new alerts were coming into the system).

We can pursue this in another thread if you like, otherwise is it ok to close this one?